Relapsed disease following conventional therapy remains one of the central problems in the treatment of leukemia. Leukemia initiating cells (LIC), possess stem cell properties such as self-renewal and pluripotency enabling them to mediate disease initiation and maintenance. Targeting the LIC compartment is therefore at the core of novel therapeutic approaches to leukemia. We have demonstrated that the Promyelocytic Leukemia gene (PML) is critical to HSC and CML LIC maintenance. Notably, pharmacologic ablation of PML using arsenic trioxide (ATO) combined with cytotoxic therapy resulted in the exhaustion of LICs and eradication of disease in a BCR-ABL CML murine model. Our preclinical findings led to a phase I trial at our institution combining TKI and ATO in CML patients with evidence of persistent disease, as well as a multi-center, randomized trial in China comparing TKI and TKI plus a novel oral arsenic in CML. Encouraging interim results of our own trial prompted the addition of a new site as well as collaboration with our colleagues in China. Our ongoing studies of PML's role in HSC and LIC maintenance has uncovered evidence of a novel non-cell- autonomous role for PML in HSCs and in CML and AML LICs through its expression in mesenchymal stem cells suggesting that PML targeting strategies could be dually effective against the leukemic niche and LIC proper. Furthermore, we discovered a cell-autonomous, metabolic dimension to PML-orchestrated HSC maintenance via the nutrient sensor, peroxisome proliferator activating receptor ? (PPAR ?) and fatty acid oxidation (FAO) which promote HSC asymmetric division. In AML, we have intriguing evidence that abrogating PPAR signaling with a novel PPAR? inhibitor markedly inhibits colony initiating capacity. In addition, our preliminary data indicates that the combination of ATO and PPAR inhibition results in synergistic cytotoxicity in part through increased ROS production. Given the ROS-low status of the AML LIC compartment, we postulate that this combination could efficiently target the LIC compartment in specific subtypes of AML, such as IDH mutant leukemia, which is known to have limited reducing capacity. To translate the potential of targeting PML pathways, we propose to conduct the following Specific Aims: (1) to define the subtypes of AML that rely on Pml expression in MSCs for LIC maintenance and dissect the MSC-derived factors that mediate LIC maintenance in CML and AML (2) to study the role of the PML/PPAR/FAO pathway in AML LICs; (3) to assess the effectiveness of combinatorial treatments targeting PML pathways, against AML LICs in preclinical murine models, including IDH2-mutant models, with evaluation of PML targeting strategies in combination with IDH2-mutant deinduction and (4) to develop a clinical trial of PPAR? inhibition alone or in combination with ATO for the treatment of relapsed, refractory AML. The studies proposed will significantly impact the treatment of CML and AML patients by targeting the reservoir of leukemic cells that mediates disease persistence and relapse.